Group_tidytext_Final
Akhil Havaldar
10/20/2021
Pacific Region
## # A tibble: 4,776 x 7
## newspapers word n total tf idf tf_idf
## <chr> <chr> <dbl> <dbl> <dbl> <dbl> <dbl>
## 1 Alaska Dispatch News alaska 26 3864 0.00673 1.10 0.00739
## 2 LATimes joshua 18 4855 0.00371 1.10 0.00407
## 3 The Spokesman-Review mammoth 16 4489 0.00356 1.10 0.00392
## 4 The Spokesman-Review simpson 14 4489 0.00312 1.10 0.00343
## 5 The Spokesman-Review bones 12 4489 0.00267 1.10 0.00294
## 6 Alaska Dispatch News alaskans 11 3864 0.00285 1.10 0.00313
## 7 Alaska Dispatch News utqiagvik 11 3864 0.00285 1.10 0.00313
## 8 The Spokesman-Review woolly 10 4489 0.00223 1.10 0.00245
## 9 Alaska Dispatch News ocean 9 3864 0.00233 1.10 0.00256
## 10 The Spokesman-Review amazon 9 4489 0.00200 1.10 0.00220
## # ... with 4,766 more rowsLA Times
Alaska Dispatch News
The Spokesman-Review
Rocky Mountain Region
## # A tibble: 4,252 x 7
## newspapers word n total tf idf tf_idf
## <chr> <chr> <dbl> <dbl> <dbl> <dbl> <dbl>
## 1 The Salt Lake Tribune s 24 5491 0.00437 1.10 0.00480
## 2 Idaho Falls Post Register yellowstone 20 3388 0.00590 1.10 0.00649
## 3 Idaho Falls Post Register idaho 12 3388 0.00354 1.10 0.00389
## 4 The Wyoming Tribune birds 11 3552 0.00310 1.10 0.00340
## 5 The Salt Lake Tribune adapting 10 5491 0.00182 1.10 0.00200
## 6 Idaho Falls Post Register houston 8 3388 0.00236 1.10 0.00259
## 7 Idaho Falls Post Register pierce 8 3388 0.00236 1.10 0.00259
## 8 Idaho Falls Post Register legislators 7 3388 0.00207 1.10 0.00227
## 9 Idaho Falls Post Register foreman 6 3388 0.00177 1.10 0.00195
## 10 Idaho Falls Post Register wrote 6 3388 0.00177 1.10 0.00195
## # ... with 4,242 more rowsIdaho Falls Post Register
The Salt Lake Tribune
The Wyoming Tribune
Midwest Region
## # A tibble: 7,335 x 7
## titles_midwest word n total tf idf tf_idf
## <chr> <chr> <dbl> <dbl> <dbl> <dbl> <dbl>
## 1 Grim forecast in Minnesota climate ada~ the 76 1451 0.0524 0 0
## 2 Climate change: when local is global. the 71 1506 0.0471 0 0
## 3 Climate change: when local is global. and 67 1506 0.0445 0 0
## 4 Climate change: call it what you will,~ the 65 1653 0.0393 0 0
## 5 Scientists offer new estimate for how ~ the 61 961 0.0635 0 0
## 6 K-State experts outline climate change~ the 56 1400 0.04 0 0
## 7 WANTED: LEADERS WILLING TO FACE CLIMAT~ and 49 1115 0.0439 0 0
## 8 Grim forecast in Minnesota climate ada~ to 48 1451 0.0331 0 0
## 9 Report: Climate change could increase ~ the 48 976 0.0492 0 0
## 10 Climate change, logging collide the 47 1103 0.0426 0 0
## # ... with 7,325 more rowsIllinois
Indiana
Iowa
Kansas
Michigan
Minnesota
Missouri
North Daktoa
Nebraska
Ohio
South Dakota
Wisconsin
Southwest Region
## # A tibble: 9,910 x 7
## titles_south word n total tf idf tf_idf
## <chr> <chr> <dbl> <dbl> <dbl> <dbl> <dbl>
## 1 Industrial Evolution the 258 4393 0.0587 0 0
## 2 Is Climate Change Denial Thawing in T~ the 163 2881 0.0566 0 0
## 3 Industrial Evolution to 129 4393 0.0294 0 0
## 4 WHAT WOULD DARWIN DO? the 123 1820 0.0676 0 0
## 5 Industrial Evolution of 120 4393 0.0273 0 0
## 6 Industrial Evolution a 102 4393 0.0232 0 0
## 7 Industrial Evolution and 99 4393 0.0225 0 0
## 8 High and Dry the 88 1628 0.0541 0 0
## 9 Is Climate Change Denial Thawing in T~ of 81 2881 0.0281 0 0
## 10 Industrial Evolution water 76 4393 0.0173 0.575 0.00995
## # ... with 9,900 more rowsArizona
New Mexico
Oklahoma
Texas
Northeast Region
## # A tibble: 7,590 x 7
## NE_articles word n total tf idf tf_idf
## <chr> <chr> <dbl> <dbl> <dbl> <dbl> <dbl>
## 1 Northeast Article 3 the 225 3434 0.0655 0 0
## 2 Northeast Article 12 the 220 2957 0.0744 0 0
## 3 Northeast Article 13 the 128 1826 0.0701 0 0
## 4 Northeast Article 3 and 118 3434 0.0344 0 0
## 5 Northeast Article 12 to 103 2957 0.0348 0 0
## 6 Northeast Article 3 of 94 3434 0.0274 0 0
## 7 Northeast Article 11 the 85 1270 0.0669 0 0
## 8 Northeast Article 3 to 84 3434 0.0245 0 0
## 9 Northeast Article 12 of 77 2957 0.0260 0 0
## 10 Northeast Article 8 the 76 1018 0.0747 0 0
## # ... with 7,580 more rowsNortheast Wordclouds
Southeast Region
## # A tibble: 6,519 x 7
## SEarticles word n total tf idf tf_idf
## <chr> <chr> <dbl> <dbl> <dbl> <dbl> <dbl>
## 1 Southeast Article 15 the 130 1985 0.0655 0 0
## 2 Southeast Article 1 the 115 1932 0.0595 0 0
## 3 Southeast Article 7 the 87 1582 0.0550 0 0
## 4 Southeast Article 15 of 69 1985 0.0348 0 0
## 5 Southeast Article 4 the 69 1151 0.0599 0 0
## 6 Southeast Article 10 the 59 890 0.0663 0 0
## 7 Southeast Article 1 of 54 1932 0.0280 0 0
## 8 Southeast Article 12 the 53 1147 0.0462 0 0
## 9 Southeast Article 13 the 53 913 0.0581 0 0
## 10 Southeast Article 15 in 51 1985 0.0257 0 0
## # ... with 6,509 more rowsSoutheast Wordclouds
Final Analysis
We found that almost all of the articles we chose had more negative words than positive when comparing the BING sentiment values. This is unsurprising, but where we see differences in region is the words that cause these negative values. For example, the word “misinformation” appeared with high frequency in the south region, compared to words like “emission” or “damage” from other regions across the country. We think this is an interesting difference between what different regions in the US might think the problem with climate change is. The south region had articles which seemed to think the problem is overreaction, misinformation, and government control. The pacific region had articles very fearful of the effects of climate change, as did the midwest region with “floods” and other extreme weather being high frequency words. The Rocky Mountain region was the only region with an even distribution of negative and positive valued words. However we did not think this was too surprising, given the political affiliation of states the newspapers were pulled from - Utah, Wyoming, and Idaho.
There was also an interesting difference between regions in terms of the scale of climate change. For example, articles pulled from the south region had words like “Americans” and “national”, where the articles from the northeast region had words such as “world”, “countries”, and “global”. High frequency words from the midwest and pacific regions also mentioned words of government and politics, but nothing specifically global. We thought it was interesting that only one article clearly conveyed a concern for the rest of the world, when climate change has been established as a global issue.
Another interesting difference between regions we found was whether regions had a stronger correlation with problems or solutions. Articles pulled from the southwest region had high frequency words like “energy” and “oil” and “emissions”, compared to “education” and “change” from the midwest articles. This seems to convey the idea that the midwest was more concerned for solutions to the existing effects of climate change, whereas the south region could be more focused on identifying the roots of the problem.
Overall, we think we have a general sense of feelings towards climate change over the different regions of the US. However, for future research, we think it would be a good idea to collect articles from every state, and not only use articles that came from big newspapers. We think we might see more of people’s true opinions on climate change if we do not primarily focus on articles coming from bigger cities.